OKAI GOOkanagan AI Growth Opportunities

AI Literacy Framework · Part 2 of 7

The workflow and first checks

The four habits as a workflow, and what to ask before any task goes near an AI tool.

Diligence, delegation, direction & discernment as a workflow

Creation diligenceIs AI use responsible here?Privacy, risks, impactsDelegationWho should do the job:me, AI or another person?DirectionContext, task, audience,constraints, examplesYou do itYourself, or with anotherperson, in your own voiceDiscernmentIs it true? Is it fair?Is it actually good?Transparency and accountability diligenceOwn every claim. Disclose how AI was used. You are the final checkpoint.AInot AIoutputredirectdraft for releaseCreation diligenceIs AI use responsible here?Privacy, risks, impactsDelegationWho should do the job:me, AI or another person?You do itYourself, or withanother person, inyour own voiceDirectionContext, task,audience, constraints,examplesDiscernmentIs it true?Is it fair?Is it actually good?Transparency andaccountability diligenceOwn every claim. Disclose howAI was used. You are thefinal checkpoint.not AIAIoutputredirectdraft for release

Try it on one example. Someone asks you for a plain-language summary of your survey findings for their newsletter. Is the data safe to put into an AI tool? Who should write it: you, the AI, or their communications person? If AI, what direction would you give? What would you check in what comes back? What would you disclose when you send it?

Think of your last AI use. Which step did you skip?

Initial diligence checks before you use AI

Ask these before any task goes near an AI tool.

1

Privacy & data protection

Whose information can go in?

What personal information are you willing to share?

2

Risks & benefits

What could go wrong, for whom, and is the gain worth it?

3

Wider impacts

Environmental cost, social costs. Who is being harmed?

Your partner’s, employer’s or instructor’s AI policy is the first check, not the last.

Privacy is the big one when you work with people’s information. Anything covered by a consent form or an ethics approval doesn’t go into a tool that isn’t approved for it. And if the rule is no AI, the workflow ends here.

Quick data security check

Your turn — the traffic-light sort

These scenarios come from a community research placement, where a student works with a partner organization. Read “partner” as your own employer, client or instructor.

Green

Go. Low risk, and the partner allows it.

Yellow

Go, only if a condition is met. Name the condition.

Red

Don’t. Say what makes it irresponsible.